QA » openQA Project

User story¶

As a reviewer, I want to skip reviewing tests which failed due to some service malfunction, to spent my working time more effectively.
Example : When SCC-proxy is down we know that all tests which trying to register via SCC-proxy will fail . Currently it force people to review hundreds of jobs failed with exactly same reason. We need to find a way how to fail tests before they even start actual execution and in exactly same way.

Acceptance criteria¶

• In case of malfunction of service which affecting group of test, they all will fail in same way and as soon as possible
• This flow is transparent to all tests giving transparent interface allowing each test to choose which services will affect this certain test behavior

Suggestions¶

• Add to each test depend job which will contains service check up and in case of failure will skip all dependent jobs ( Cons: over complicate setup because we need to add plenty duplicate jobs to cover different combinations of version/arch/flavor
• Inject this functionality into openQA backend ( pre-run hooks based on certain job settings exposing dependent services ? )

Further details¶